Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    instances of the word "of" when it directly precedes numbers or certain phrases

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating rounds or phases in competitions or processes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hyde
    -0.79
    runtime
    -0.67
    arily
    -0.66
    sense
    -0.66
    ãĤ¼
    -0.66
    esse
    -0.65
     sqor
    -0.64
    oir
    -0.63
    ghan
    -0.63
    bh
    -0.62
    POSITIVE LOGITS
     applause
    0.99
     Playoffs
    0.79
     scrimmage
    0.79
     chemotherapy
    0.78
     eviction
    0.76
     reckoning
    0.72
     negotiations
    0.72
     competition
    0.70
     playoffs
    0.68
     hostilities
    0.67
    Act Density 0.077%

    No Known Activations